Reasoning on BIG-Bench Extra Hard
37.8ScoreQwen3-30B-A3B-Inst-2507
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen3-30B-A3B-Inst-25072026.02 | 37.8 | 1 | |
| LLaDA2.1-flashInference Mode=Q Mode2026.02 | 35.77 | 3.17 | |
| LLaDA2.1-flashInference Mode=S Mode2026.02 | 33.51 | 5.04 | |
| LLaDA2.0-flash2026.02 | 27.86 | 4.6 | |
| Ling-flash-2.02026.02 | 23.24 | 1 | |
| Qwen3-8Bno think=true2026.02 | 18.27 | — | |
| LLaDA2.0-mini2026.02 | 16.47 | 2.03 | |
| LLaDA2.1-minimode=Q Mode2026.02 | 15.78 | 1.66 | |
| LLaDA2.1-minimode=S Mode2026.02 | 15.3 | 3.19 | |
| Ling-mini-2.02026.02 | 14.81 | — |